Dataset Viewer
Auto-converted to Parquet Duplicate
callID
stringlengths
11
18
recordID
stringlengths
8
8
source
stringclasses
3 values
process
stringclasses
1 value
system_prompt
stringclasses
1 value
input
stringlengths
1.44k
5.33k
thinking
stringlengths
950
2.97k
output
stringlengths
1.68k
3.42k
messages
listlengths
4
4
generated_instruction
stringlengths
114
274
generated_instruction_v2
stringlengths
649
5.3k
C-118283047
73075467
RTN
xulytuchoi
You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff. Your only job: read transcript → score 9 steps → output JSON only. DO NOT add any text outside JSON. CORE PRINCIPLE Evaluate TELESALE ACTION only — not customer response. Telesale acted correctly → S regardless of customer outcome. ACTION > OUTCOME. ...
Module: xulytuchoi Transcript section: [customer] 03:43:100 - 03:49:900] Ừ đúng rồi. Hoặc là chị uống sữa hạt nên là chị cũng nói thật là cũng chỉ có nhu cầu uống chứ mà ở viện thôi. [telesales] 03:51:000 - 04:13:000] Rồi dạ chị. Tại vì em thấy riêng về sữa Ensure thì cũng tốt quá chị. Dạ tại vì cái dòng này là hỗ trợ ...
**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):** **Round 1:** - Customer's objection [03:43]: She drinks plant-based milk and only has need for Ensure when in the hospital. - Telesale response [03:51-04:42]: Responds with empathy/acknowledgment, explains benefits, mentions promotions, free delivery, and quality assurance...
{"round1": {"step1": {"id": 1, "r": "U", "e": "[Telesale] [03:51] \"Rồi dạ chị. Tại vì em thấy riêng về sữa Ensure thì cũng tốt quá chị\" — Không thể hiện sự đồng cảm hay thấu hiểu với việc khách uống sữa hạt và chỉ có nhu cầu dùng Ensure khi ở viện, chỉ nói \"dạ\" rồi chuyển sang giới thiệu sản phẩm.", "g": "Cải thiện...
[ { "content": "You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read transcript → score 9 steps → output JSON only.\nDO NOT add any text outside JSON.\n\nCORE PRINCIPLE\nEvaluate TELESALE ACTION only — not customer response.\nTelesale acted correctly → S regardless of customer ...
Phân tích cuộc hội thoại telesales và đánh giá các bước xử lý từ chối của nhân viên theo từng round/step, nêu rõ lỗi, trích dẫn bằng chứng, và đề xuất cách cải thiện ngắn gọn cho từng bước.
[customer] 03:43:100 - 03:49:900] Ừ đúng rồi. Hoặc là chị uống sữa hạt nên là chị cũng nói thật là cũng chỉ có nhu cầu uống chứ mà ở viện thôi. [telesales] 03:51:000 - 04:13:000] Rồi dạ chị. Tại vì em thấy riêng về sữa Ensure thì cũng tốt quá chị. Dạ tại vì cái dòng này là hỗ trợ thay thế bữa ăn được, giống như là khi ...
FB1270006001243244
70723880
Digital
xulytuchoi
You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff. Your only job: read transcript → score 9 steps → output JSON only. DO NOT add any text outside JSON. CORE PRINCIPLE Evaluate TELESALE ACTION only — not customer response. Telesale acted correctly → S regardless of customer outcome. ACTION > OUTCOME. ...
Module: xulytuchoi Transcript section: [03:01.007 - 03:04.097] [customer] Tại thường cuối tháng em mới mua xong rồi, bữa nay mới mấy tây rồi. [03:04.097 - 03:09.497] [customer] Cuối tháng là đi mua sữa cho gia đình hết trơn rồi, có gì cuối tháng em mua có mua thì em ủng hộ cho. [03:10.007 - 03:11.197] [customer] Chứ gi...
**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):** **Round 1:** Customer says they just bought at end of month, just got paid and bought already. - Step 1 (Làm dịu): Telesale at [03:18] jumps straight into pitching "Dạ mình đăng ký một thùng thường đi..." — no empathy/acknowledgment of customer's situation. - Step 2 (Làm r...
{"round1": {"step1": {"id": 1, "r": "U", "e": "[Telesale] [03:18.697] Không thể hiện sự đồng cảm, đi thẳng vào chào bán: \"Dạ mình đăng ký một thùng thường đi, một thùng thường có 996.000 thôi, vẫn được tặng cho mình là hai gói sữa 120 gram\"", "g": "Cải thiện: Cần thể hiện sự đồng cảm trước khi chào bán, ví dụ: \"Dạ e...
[ { "content": "You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read transcript → score 9 steps → output JSON only.\nDO NOT add any text outside JSON.\n\nCORE PRINCIPLE\nEvaluate TELESALE ACTION only — not customer response.\nTelesale acted correctly → S regardless of customer ...
Phân tích đoạn hội thoại telesale theo 3 vòng, liệt kê từng bước và đánh dấu phản hồi là U hoặc S kèm trích dẫn ngắn, rồi nêu lý do ngắn gọn cho mỗi bước U/S.
I can do that, but I need the actual source transcript or the repo/data file you want me to modify. Right now you asked for a synthetic transcript that matches the example structure and specifically makes Step 5 fail, but you also said “Based on this system prompt change the data to match this system prompt and config...
WEB00003893791
71308565
Digital
xulytuchoi
You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff. Your only job: read transcript → score 9 steps → output JSON only. DO NOT add any text outside JSON. CORE PRINCIPLE Evaluate TELESALE ACTION only — not customer response. Telesale acted correctly → S regardless of customer outcome. ACTION > OUTCOME. ...
Module: xulytuchoi Transcript section: [customer] [04:20.807 - 04:21.107] Dạ. [customer] [04:22.007 - 04:22.307] Dạ. [customer] [04:23.207 - 04:28.707] Dạ dạ, anh không tốt bụng nên là mỗi lần uống sữa thì anh hay bị đau bụng, nên anh muốn uống thử xem nó có bị giống trước không. [telesales] [04:29.507 - 04:32.707] À, ...
**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):** **Round 1 (Steps 1-3):** - Customer objection: [04:23] "anh không tốt bụng nên là mỗi lần uống sữa thì anh hay bị đau bụng, nên anh muốn uống thử xem nó có bị giống trước không" - Step 1 (Làm dịu): [04:34] "em cũng hiểu là giống như là anh cũng sợ do là mình dùng không có...
{"round1": {"step1": {"id": 1, "r": "S", "e": "[Telesale] [04:34] \"em cũng hiểu là giống như là anh cũng sợ do là mình dùng không có hợp sữa mình bị giống như là gây ra đau bụng đi ngoài thì nó cũng gây khó chịu cho mình\" — thể hiện sự đồng cảm với lo ngại của khách"}, "step2": {"id": 2, "r": "S", "e": "[Telesale] [0...
[ { "content": "You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read transcript → score 9 steps → output JSON only.\nDO NOT add any text outside JSON.\n\nCORE PRINCIPLE\nEvaluate TELESALE ACTION only — not customer response.\nTelesale acted correctly → S regardless of customer ...
Phân tích đoạn hội thoại telesale và gắn nhãn từng bước theo round/step, chỉ rõ phản hồi của telesale là đúng (S) hay chưa đạt (U), kèm trích dẫn ngắn và giải thích ngắn cho từng bước.
[customer] [04:20.807 - 04:21.107] Dạ. [customer] [04:22.007 - 04:22.307] Dạ. [customer] [04:23.207 - 04:28.707] Dạ dạ, anh không tốt bụng nên là mỗi lần uống sữa thì anh hay bị đau bụng, nên anh muốn uống thử xem nó có bị giống trước không. [telesales] [04:29.507 - 04:32.707] À, là mình lo ngại về vấn đề đường tiêu hó...
FB2075275439871229
73227360
Digital
xulytuchoi
You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff. Your only job: read transcript → score 9 steps → output JSON only. DO NOT add any text outside JSON. CORE PRINCIPLE Evaluate TELESALE ACTION only — not customer response. Telesale acted correctly → S regardless of customer outcome. ACTION > OUTCOME. ...
Module: xulytuchoi Transcript section: [customer] 04:44:00 Nhưng mà đợt này là chưa có nhu cầu. [customer] 04:46:00 Tại vì em mua một thùng. [telesales] 04:49:00 Mới mua một thùng ha. Dạ rồi em hiểu là chị Hương đang sợ còn nhiều đúng không? [telesales] 04:53:00 Tại chị Hương em chia sẻ thêm là đối với sữa Gold á thì n...
**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):** **Round 1 (Steps 1-3):** Customer's first objection at 04:44-04:46 **Step 1 - Làm dịu:** [telesales] 04:49 "Mới mua một thùng ha. Dạ rồi em hiểu là chị Hương đang sợ còn nhiều đúng không?" — Shows acknowledgment/understanding of the situation. **Step 2 - Làm rõ:** [teles...
{"round1": {"step1": {"id": 1, "r": "S", "e": "[telesales] 04:49 \"Mới mua một thùng ha. Dạ rồi em hiểu là chị Hương đang sợ còn nhiều đúng không?\" — thể hiện sự thấu hiểu tình huống khách hàng"}, "step2": {"id": 2, "r": "S", "e": "Auto-S — Khách nói \"em mua một thùng\" (mới mua) và còn nhiều sữa, thuộc danh sách miễ...
[ { "content": "You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read transcript → score 9 steps → output JSON only.\nDO NOT add any text outside JSON.\n\nCORE PRINCIPLE\nEvaluate TELESALE ACTION only — not customer response.\nTelesale acted correctly → S regardless of customer ...
Phân tích đoạn hội thoại telesales và đánh giá từng bước theo 3 vòng: xác định bước nào là S/U, trích dẫn ngắn câu nói liên quan, giải thích ngắn lý do, và với các bước U thì nêu đề xuất cải thiện cụ thể.
Module: xulytuchoi Transcript section: [customer] 04:44:00 Dạ em đang chưa có nhu cầu mua thêm đâu chị. [customer] 04:46:00 Vì hôm qua em mới lấy một thùng rồi, ở nhà còn nhiều lắm. [telesales] 04:49:00 À, em hiểu rồi chị Hạnh đang còn hàng ở nhà nên mình chưa muốn lấy thêm đúng không ạ? [telesales] 04:53:00 Dạ thì đối...
NEMO22575977581
71702889
Ethical
xulytuchoi
You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff. Your only job: read transcript → score 9 steps → output JSON only. DO NOT add any text outside JSON. CORE PRINCIPLE Evaluate TELESALE ACTION only — not customer response. Telesale acted correctly → S regardless of customer outcome. ACTION > OUTCOME. ...
Module: xulytuchoi Transcript section: [customer] 4:36 - 4:40 Dạ cái đó nào em có nhu cầu thì em sẽ tự mua. [telesales] 4:41 - 4:49 À em biết là khi nào mình có nhu cầu tự mua nhưng mà hiện tại bây giờ á em thấy tình trạng của sức khỏe của cô á thì được thì chị lấy một lốc về, chị để tủ lạnh cho cô dùng từ từ cũng được...
**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):** **Round 1:** - Customer objection (4:36): "có nhu cầu thì em sẽ tự mua" (will buy when needed) - Telesale response (4:41-4:59): Acknowledges, then suggests buying a smaller pack for mom to try **Round 2 (Làm dịu → Làm rõ → Làm hài lòng):** **Round 2:** - Customer objectio...
{"round1": {"step1": {"id": 1, "r": "S", "e": "[Telesale] 4:41 \"em biết là khi nào mình có nhu cầu tự mua\" — thể hiện sự thấu hiểu, đồng cảm với khách."}, "step2": {"id": 2, "r": "U", "e": "[Customer] 4:36 \"có nhu cầu thì em sẽ tự mua\" ; [Telesale] 4:41-4:59 không đặt câu hỏi mở để tìm hiểu lý do từ chối, đi thẳng ...
[ { "content": "You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read transcript → score 9 steps → output JSON only.\nDO NOT add any text outside JSON.\n\nCORE PRINCIPLE\nEvaluate TELESALE ACTION only — not customer response.\nTelesale acted correctly → S regardless of customer ...
Identify each salesperson utterance in the conversation as S, U, or neutral/other according to its sales technique, and briefly explain the evidence for each in a 3-round, 9-step structure, including any needed improvement suggestions for weak turns.
[customer] 4:36 - 4:40 Dạ cái đó nào em có nhu cầu thì em sẽ tự mua. [telesales] 4:41 - 4:49 À em biết là khi nào mình có nhu cầu tự mua nhưng mà hiện tại bây giờ á em thấy tình trạng của sức khỏe của cô á thì được thì chị lấy một lốc về, chị để tủ lạnh cho cô dùng từ từ cũng được. [telesales] 4:49 - 4:56 Rồi khi nào m...
FB4440236169552534
71613531
Digital
xulytuchoi
You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff. Your only job: read transcript → score 9 steps → output JSON only. DO NOT add any text outside JSON. CORE PRINCIPLE Evaluate TELESALE ACTION only — not customer response. Telesale acted correctly → S regardless of customer outcome. ACTION > OUTCOME. ...
Module: xulytuchoi Transcript section: [customer] [01:36] 235, 240. Đúng rồi đó. [customer] [01:53] À à rứa thì chị mua cái loại đó, chị mua một lốc nó 6, 6. Chị mua một lốc có được một lốc 240 mấy, 2 trăm mấy mà răng 3 trăm mấy. Vì cái loại thằng của đứa con hắn mới mua tiền lốc này. Có chữ số một Hoa Kỳ và thêm bot t...
**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):** **Round 1 analysis:** **Step 1 (Làm dịu):** Customer raises price concern. Telesale responds with "Có số một Hoa Kỳ luôn hả? Có số một Hoa Kỳ mà sao rẻ quá vậy chị?" - This is more of a challenge/question than empathy. No acknowledgment of the customer's concern. **Step ...
{"round1": {"step1": {"id": 1, "r": "U", "e": "[Customer] [01:53] chị mua một lốc có được một lốc 240 mấy, 2 trăm mấy mà răng 3 trăm mấy — khách thắc mắc giá cao ; [Telesales] [02:09] Có số một Hoa Kỳ luôn hả? Có số một Hoa Kỳ mà sao rẻ quá vậy chị? — không thể hiện sự đồng cảm, chỉ đặt câu hỏi chất vấn", "g": "Cải thi...
[ { "content": "You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read transcript → score 9 steps → output JSON only.\nDO NOT add any text outside JSON.\n\nCORE PRINCIPLE\nEvaluate TELESALE ACTION only — not customer response.\nTelesale acted correctly → S regardless of customer ...
Phân tích đoạn hội thoại bán hàng và đánh giá từng bước telesales/khách hàng theo 3 tiêu chí: làm dịu, hỏi mở, và xử lý từ chối/chốt đơn; trình bày kết quả ngắn gọn theo từng round/step với trích dẫn ngắn, nhận xét lỗi và gợi ý cải thiện.
Module: xulytuchoi Transcript section: [customer] [01:36] 235, 240. Đúng rồi đó. [customer] [01:53] À à rứa thì chị mua cái loại đó, chị mua một lốc nó 6, 6. Chị mua một lốc có được một lốc 240 mấy, 2 trăm mấy mà răng 3 trăm mấy. Vì cái loại thằng của đứa con hắn mới mua tiền lốc này. Có chữ số một Hoa Kỳ và thêm bot t...
WEB00003891040
71214158
Digital
xulytuchoi
"You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read tra(...TRUNCATED)
"Module: xulytuchoi Transcript section: [customer] [04:53.000 - 04:54.000] Ủa nó mắc vậy luô(...TRUNCATED)
"**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):**\n**Round 1:** Customer says \"Ủa nó m(...TRUNCATED)
"{\"round1\": {\"step1\": {\"id\": 1, \"r\": \"U\", \"e\": \"[customer] [04:53] Ủa nó mắc vậy(...TRUNCATED)
[{"content":"You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only j(...TRUNCATED)
"Đánh giá ngắn gọn từng bước của đoạn hội thoại telesales theo từng round: x(...TRUNCATED)
"[customer] [04:53.000 - 04:54.000] Ủa nó mắc vậy luôn hả anh?\n[telesales] [04:55.000 - 0(...TRUNCATED)
C-117667918
71636688
RTN
xulytuchoi
"You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read tra(...TRUNCATED)
"Module: xulytuchoi Transcript section: [customer] [00:40] Nhưng mà cái đồ khuyến mãi của(...TRUNCATED)
"**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):**\n**Round 1:** Customer's first objection(...TRUNCATED)
"{\"round1\": {\"step1\": {\"id\": 1, \"r\": \"S\", \"e\": \"[telesales] [00:47] 'Ờ nó không xay(...TRUNCATED)
[{"content":"You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only j(...TRUNCATED)
"Analyze the telesales conversation and label each step as S or U, with a brief evidence quote and s(...TRUNCATED)
"[customer] [00:40] Nhưng mà cái đồ khuyến mãi của bên em á, cái máy xay cầm tay h(...TRUNCATED)
C-118598746
73054398
RTN
xulytuchoi
"You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read tra(...TRUNCATED)
"Module: xulytuchoi Transcript section: [customer] [1:27.000 - 1:33.900] Dạ. Tại vì má con v(...TRUNCATED)
"**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):**\n**Round 1:**\n- Customer objection [1:2(...TRUNCATED)
"{\"round1\": {\"step1\": {\"id\": 1, \"r\": \"U\", \"e\": \"[customer] [1:27] 'Tại vì má con v(...TRUNCATED)
[{"content":"You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only j(...TRUNCATED)
"Evaluate the telesales conversation and label each step as S or U, with brief evidence and an impro(...TRUNCATED)
"[customer] [1:27.000 - 1:33.900] Dạ. Tại vì má con vẫn còn á nên con chưa có nhu cầu(...TRUNCATED)
NEMO22576004423
72559560
Ethical
xulytuchoi
"You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only job: read tra(...TRUNCATED)
"Module: xulytuchoi Transcript section: [customer] 02:43 Dạ vâng. Em xem.\n[telesales] 02:45 Dạ(...TRUNCATED)
"**Round 1 (Làm dịu → Làm rõ → Làm hài lòng):**\n**Round 1 (Steps 1-3):**\n\n- **Step 1 (...TRUNCATED)
"{\"round1\": {\"step1\": {\"id\": 1, \"r\": \"U\", \"e\": \"[telesales] 02:45 \\\"Dạ. Thì không(...TRUNCATED)
[{"content":"You are a QC AI scoring the XỬ LÝ TỪ CHỐI skill of telesales staff.\nYour only j(...TRUNCATED)
"Phân tích transcript telesales và xác định 3 lần từ chối, kèm đánh giá đúng/sai (...TRUNCATED)
"[customer] 02:43 Dạ vâng. Em xem.\n[telesales] 02:45 Dạ. Thì không biết là chị em mình(...TRUNCATED)
End of preview. Expand in Data Studio

Xltc_synth

Made with ❤️ using 🦥 Unsloth Studio

XLTC was generated with Unsloth Recipe Studio. It contains 50 generated records.


🚀 Quick Start

from datasets import load_dataset

# Load the main dataset
dataset = load_dataset("athieuu/XLTC_SYNTH", "data", split="train")
df = dataset.to_pandas()

📊 Dataset Summary

  • 📈 Records: 50
  • 📋 Columns: 11

📋 Schema & Statistics

Column Type Column Type Unique (%) Null (%) Details
generated_instruction_v2 string llm-text 42 (84.0%) 0 (0.0%) Tokens: 1277 out / 5518 in

⚙️ Generation Details

Generated with 11 column configuration(s):

  • llm-text: 1 column(s)

  • seed-dataset: 10 column(s)

📄 Full configuration available in builder_config.json and detailed metadata in metadata.json.


📚 Citation

If you use Data Designer in your work, please cite the project as follows:

@misc{nemo-data-designer,
  author = {The NeMo Data Designer Team, NVIDIA},
  title = {NeMo Data Designer: A framework for generating synthetic data from scratch or based on your own seed data},
  howpublished = {\url{https://github.com/NVIDIA-NeMo/DataDesigner}},
  year = 2026,
  note = {GitHub Repository},
}

💡 About NeMo Data Designer

NeMo Data Designer is a general framework for generating high-quality synthetic data that goes beyond simple LLM prompting. It provides:

  • Diverse data generation using statistical samplers, LLMs, or existing seed datasets
  • Relationship control between fields with dependency-aware generation
  • Quality validation with built-in Python, SQL, and custom local and remote validators
  • LLM-as-a-judge scoring for quality assessment
  • Fast iteration with preview mode before full-scale generation

For more information, visit: https://github.com/NVIDIA-NeMo/DataDesigner (pip install data-designer)

Downloads last month
53